Adoptive immunotherapy with genetically modified T cells expressing a chimeric antigen receptor (CAR) is highly promising for B-cell malignancies. However, fewer than half of patients with relapsed non-Hodgkin lymphoma (NHL) achieve durable remissions following treatment with CD19-targeted CAR T cells. In some cases this results from target antigen loss or rejection of the infused cells due to immunogenic murine CAR components, but in most cases the causes of treatment resistance or relapse after an initial response remain poorly understood. We propose a plan to elucidate the reasons for treatment failure in a recently initiated CAR T cell clinical trial by carefully evaluating biological features of the tumor and tumor microenvironment before and after CAR T cell therapy as well as phenotypes of patient T cells and infused CAR T cells. As part of our phase I/II clinical trial of a fully human 3rd generation CD20-specific CAR in patients with relapsed or refractory B-cell NHL (funded by a separate source), all patients undergo mandatory tumor biopsies before and after treatment. This will allow us to discover biological characteristics predictive of responsiveness to treatment, and to evaluate adaptive changes in the tumor over time to reveal the mechanisms of immune escape leading to relapse. We will employ a step-wise approach using state-of- the-art methodologies, including multicolor flow cytometry, single-cell RNA sequencing and gene expression profiling, and multiplex immunohistochemistry. We have assembled a world class team of investigators that will evaluate the potential obstacles to successful therapy, including tumor entry barriers, tumor infiltration by suppressive cells, CAR T-cell exposure to inhibitory ligands or secreted proteins, and CAR T-cell exhaustion. There is a robust body of preclinical data demonstrating that less-differentiated T cell subsets impart superior in vivo expansion, persistence, and anti-tumor efficacy, compared with more differentiated T cell subsets. Recent data from a small trial suggests that the frequency of these less-differentiated CD8+ cell subsets before and after CAR T cell manufacturing correlates with clinical responses in patients with chronic lymphocytic lymphoma receiving CD19-targeted CAR T cells. These results have important implications, but must be validated in other settings. We will quantify less-differentiated T cell subtypes prior to leukapheresis, as well as in the infused CAR T cell products, and correlate these characteristics with anti-tumor responses and in vivo expansion and persistence. We anticipate that these correlative studies will yield critical insights into the reasons why CAR T cell therapy is successful for some NHL patients but not others. We are hopeful that our findings will help to guide patient selection and counseling, and inform future strategies to overcome these obstacles through improved cell manufacturing technologies, CAR vector design, and/or combinatorial adjuvant therapies, not only for CD20-specific CAR T cells, but also for CAR T cell therapy for other targets.